Ignoring e2e tests for validate building and installing pipeline actions - second try + some test helpers improvements

Author: petyaslavova

.github/workflows/install_and_test.sh

@@ -40,7 +40,7 @@ cd ${TESTDIR}
 # install, run tests
 pip install ${PKG}
 # Redis tests
-pytest -m 'not onlycluster'
+pytest -m 'not onlycluster' --ignore=tests/test_scenario
 # RedisCluster tests
 CLUSTER_URL="redis://localhost:16379/0"
 CLUSTER_SSL_URL="rediss://localhost:27379/0"

redis/_parsers/base.py

@@ -191,7 +191,7 @@ def parse_moving_msg(response):
         # Expected message format is: MOVING <seq_number> <time> <endpoint>
         id = response[1]
         ttl = response[2]
-        if response[3] in [b"null", "null"]:
+        if response[3] is None:
             host, port = None, None
         else:
             value = response[3]

tests/test_maintenance_events_handling.py

@@ -237,7 +237,7 @@ def send(self, data):
                 # Format: >4\r\n$6\r\nMOVING\r\n:1\r\n:1\r\n+null\r\n (4 elements: MOVING, id, ttl, null)
                 # Note: Using + instead of $ to send as simple string instead of bulk string
                 moving_push = (
-                    f">4\r\n$6\r\nMOVING\r\n:1\r\n:{MOVING_TIMEOUT}\r\n+null\r\n"
+                    f">4\r\n$6\r\nMOVING\r\n:1\r\n:{MOVING_TIMEOUT}\r\n_\r\n"
                 )
                 response = moving_push.encode() + response
             elif b"key_receive_moving_" in data:

tests/test_scenario/conftest.py

@@ -14,11 +14,13 @@
 RELAX_TIMEOUT = 30
 CLIENT_TIMEOUT = 5
 
+DEFAULT_ENDPOINT_NAME = "m-standard"
+
 
 @pytest.fixture()
 def endpoint_name(request):
     return request.config.getoption("--endpoint-name") or os.getenv(
-        "REDIS_ENDPOINT_NAME", "m-standard"
+        "REDIS_ENDPOINT_NAME", DEFAULT_ENDPOINT_NAME
     )
 
 
@@ -75,6 +77,8 @@ def _get_client_maint_events(
     host = parsed.hostname
     port = parsed.port
 
+    tls_enabled = True if parsed.scheme == "rediss" else False
+
     if not host:
         raise ValueError(f"Could not parse host from endpoint URL: {endpoints[0]}")
 
@@ -101,6 +105,9 @@ def _get_client_maint_events(
         socket_timeout=CLIENT_TIMEOUT if socket_timeout is None else socket_timeout,
         username=username,
         password=password,
+        ssl=tls_enabled,
+        ssl_cert_reqs="none",
+        ssl_check_hostname=False,
         protocol=3,  # RESP3 required for push notifications
         maintenance_events_config=maintenance_config,
         retry=retry,

tests/test_scenario/hitless_upgrade_helpers.py

@@ -220,6 +220,7 @@ def find_target_node_and_empty_node(
     def find_endpoint_for_bind(
         fault_injector: FaultInjectorClient,
         endpoint_config: Dict[str, Any],
+        endpoint_name: str,
         timeout: int = 60,
     ) -> str:
         """Find the endpoint ID from cluster status.
@@ -252,13 +253,13 @@ def find_endpoint_for_bind(
             elif endpoints_section_started and line and not line.startswith("DB:ID"):
                 # Parse endpoint line: db:1  m-standard  endpoint:1:1  node:2  single  No
                 parts = line.split()
-                if len(parts) >= 3:
+                if len(parts) >= 3 and parts[1] == endpoint_name:
                     endpoint_full = parts[2]  # endpoint:1:1
                     if endpoint_full.startswith("endpoint:"):
                         endpoint_id = endpoint_full.replace("endpoint:", "")  # 1:1
                         return endpoint_id
 
-        raise ValueError("No endpoint ID found in cluster status")
+        raise ValueError(f"No endpoint ID for {endpoint_name} found in cluster status")
 
     @staticmethod
     def execute_rladmin_migrate(

tests/test_scenario/test_hitless_upgrade.py

@@ -48,6 +48,7 @@ def setup_and_cleanup(
         client_maint_events: Redis,
         fault_injector_client: FaultInjectorClient,
         endpoints_config: Dict[str, Any],
+        endpoint_name: str,
     ):
         # Initialize cleanup flags first to ensure they exist even if setup fails
         self._migration_executed = False
@@ -70,7 +71,7 @@ def setup_and_cleanup(
 
         try:
             self.endpoint_id = ClusterOperations.find_endpoint_for_bind(
-                fault_injector_client, endpoints_config
+                fault_injector_client, endpoints_config, endpoint_name
             )
             logging.info(f"Using endpoint: {self.endpoint_id}")
         except Exception as e:
@@ -396,7 +397,7 @@ def test_timeout_handling_during_migrating_and_moving(
 
         """
         logging.info(f"Testing timeout handling for endpoint type: {endpoint_type}")
-        client = _get_client_maint_events(endpoints_config, endpoint_type)
+        client = _get_client_maint_events(endpoints_config=endpoints_config, endpoint_type=endpoint_type)
 
         # Create three connections in the pool
         logging.info("Creating three connections in the pool.")
@@ -501,7 +502,7 @@ def test_new_connection_handling_during_migrating_and_moving(
         endpoints_config: Dict[str, Any],
     ):
         logging.info(f"Testing timeout handling for endpoint type: {endpoint_type}")
-        client = _get_client_maint_events(endpoints_config, endpoint_type)
+        client = _get_client_maint_events(endpoints_config=endpoints_config, endpoint_type=endpoint_type)
 
         logging.info("Creating one connection in the pool.")
         first_conn = client.connection_pool.get_connection()
@@ -614,7 +615,7 @@ def test_disabled_handling_during_migrating_and_moving(
     ):
         logging.info("Creating client with disabled notifications.")
         client = _get_client_maint_events(
-            endpoints_config,
+            endpoints_config=endpoints_config,
             enable_maintenance_events=False,
         )
 
@@ -741,7 +742,7 @@ def test_command_execution_during_migrating_and_moving(
         socket_timeout = 0.5
 
         client = _get_client_maint_events(
-            endpoints_config,
+            endpoints_config=endpoints_config,
             endpoint_type=endpoint_type,
             disable_retries=True,
             socket_timeout=socket_timeout,